Publications (13) View all
-
Conference Proceeding: Testing Single Frequency GPS Receivers Under Ionospheric Disturbances: A New Approach
[show abstract] [hide abstract]
ABSTRACT: The assessment of the robustness of GNSS receivers against ionosphere perturbations can be achieved in two ways: by testing the receivers in actual environments, through measurement campaigns, or by testing them in laboratory using RF constellation simulators. In actual environments, receivers are stressed not only by ionospheric disturbances but also by other perturbations such as thermal noise, multi-path or interferences. As a consequence, it makes it difficult to evaluate the performance of the receivers against the ionosphere disturbances only. In addition it is impossible to replay a particular ionospheric event to test several receivers under the same conditions. An interesting alternative to field testing is to test receivers in the laboratory using GNSS RF simulators. Indeed, GNSS constellation simulators provide the capability to create a completely controlled environment, in which each propagation effects can be isolated, and in addition they allow to repeat the same scenarios to test different receivers in the same environment. The number of tests being conducted in laboratory can be high, unlike field testing that can be expensive, which permit to extract significant statistical results. The representativeness of the disturbances being reproduced is of prime importance and special care must be taken in order to be as close as possible to the real perturbations. The methods presented here explain how to use RF simulators to assess the behavior of GNSS receivers with regard to ionospheric delays. The study was limited to both GPS L1 C/A stand-alone and EGNOS enabled receivers. Simulations are conducted using a Spirent STR4760 RF signal generator. This simulator is able to apply ionospheric error delays on simulated signals and allows to broadcast custom corrections on navigation messages. The ionospheric correction models used by the receiver can then be either the 8 coefficients of the Klobuchar model (transmitted on GPS L1 navigation message) or the ionospheric grid correction transmitted by EGNOS. The ionospehric delay time series applied to the signal are computed based on real measurements made by GNSS reference stations, for example from the IGS network of stations. A dual frequency algorithm is used to extract from the RINEX observation data the ionospheric delay component that affect the received signal. That delay is then replayed in the simulator. The broadcasted GPS navigation message Klobuchar correction coefficients can be downloaded from the IGS servers (broadcast data) and the EGNOS corrections can be found in navigation messages available on the EGNOS Message Servers. The use of real measurements to extract the ionospheric delays insures a very good representativeness with regards to the real ionospheric delays. The simulated scenarios are constructed to be coherent with the ionospheric delay data obtained from the reference measurement: they will correspond to the same date, the same constellation configuration, and the ionospheric corrections broadcasted in the navigation messages will be the same as the ones broadcasted by the real satellites. The drawback of this method is however that the data set obtained is valid only in a limited range around the position of the GNSS reference station used to compute ionospheric delay data, which limits the choice of the constellation configuration, date and positions. In spite of these limitations, relevant tests can still be done to evaluate receivers’ performance and to measure the impact of ionospheric propagation. In order to bypass the limitations of previous method, a solution would be to use a generic ionospheric model, such as the NeQuick model, to compute the ionospheric delays. In that case no limitation of date nor position exists. However it would imply to be able to reconstruct the associated correction data to be broadcasted in the navigation messages. Some methods have been developed to build the corrections models (Klobuchar and EGNOS) from the simulated delays time series, and they will be presented in this paper. After scenarios are defined, the tests are performed using a Receiver Test Bench that allows the automatic execution of a high number of tests. Scenarios are loaded automatically in the RF signal generator and the receiver outputs are collected and analyzed by a dedicated software. The accuracy of all the receivers tested can be evaluated and compared in different conditions: firstly, without any ionospheric degradations, which would represent an “ideal” case and provides the positioning accuracy of the receivers without ionosphere impairments; secondly, with ionospheric effects but without any ionospheric correction models, which represents the worst case scenario and allows to evaluate the effect of ionosphere on the degradation of the position calculation; and finally with ionospheric delays and broadcasted corrections to evaluate the gain brought by different correction models. This procedure can be used to compare the effects of various ionospheric activities (normal or exceptional) and to compare the effectiveness of different correction models. The paper will finally present some results that have been obtained for the two described methods and for different receivers under test.EuCAP, Rome (Italy); 01/2011 -
Article: Degraded Modes Resulting From The Multi-Constellation Use Of GNSS
Thèse en ligne Institut National Polytechnique. 01/2010; -
Thesis: Degraded Modes Resulting From the Multiconstellation Use Of GNSS
01/2010, Degree: PhD, Supervisor: Christophe Macabiau -
Conference Proceeding: Performance assessment of a Juzzle-based GNSS Simulator
ENC GNSS, Napoli; 05/2009 -
Conference Proceeding: Development Of A Flexible Real Time GNSS Software Receiver
ENC GNSS; 05/2009
